The verification of call center customers based on voice data as a part of multi-factor authentication for security improvement.
The use of deep neural architecture for comparing voice prints from a database, with verification accuracy of over 90%.
Creation and training of an effective ML-model using a deep neural architecture that can compare incoming voice data against voice prints from the previously generated database.
Improved authentication system security; reduced time for processing and comparing voice prints; enhanced estimated accuracy of verification.
Low EER (Exact Error Rate)
Request full cases: